Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

Earn the coveted Fabric Analytics Engineer certification. 100% off your exam for a limited time only!

Reply
ccarpent
Helper IV
Helper IV

Data Model

Hi

 

Data Model; fact and dimesnion tables, best approach best practice.

 

My datsource is a combination of SAP BI Suit, Excel and SQL views, which tend to come with a good number of columns.  More like a large mass of rows and columns; I say a mass of data but in the scheme of thinngs its not many rows 27k for example.   What I am trying to do is break this down into smaller manageable chunks or tables.  I am trying to identify what parts of my data can be tuned into either FACT or DIMENSION; is this the correct approach to take or do people just work off a flat file and thats it?

 

For example, what I mean is,  I would extract all the Supplier details from my datasourse and then turn this into a seperate table called SupplierDetails, also creating a relation using the SupplierID.  Again from the same table, I would extract all the Manager names and roles, again turn this in to a seperate table ussign a ManagerID.  My aim is then to link or create a realtionships amongst these tables to make the data more manageable.

 

Is this the wrong approach?

 

Any advice or links welcome.

 

Thanks again

Chris

 

1 ACCEPTED SOLUTION
EricHulshof
Solution Sage
Solution Sage

It is the perfect approach. I know it can be hard to determain Fact and Dim, struggled with it just a week ago. But yes, star scheme's work the best in PBI.

Also dont forget a date table (dim) 😉

EDIT:
Besides that, if you really have alot of columns you can manage them and put them in maps. I allways do that with measure and dates and that cleans up alot. example:
Knipsel.PNG Knipsel2.PNG


Quality over Quantity


Did I answer your question? Mark my post as a solution!


View solution in original post

2 REPLIES 2
EricHulshof
Solution Sage
Solution Sage

It is the perfect approach. I know it can be hard to determain Fact and Dim, struggled with it just a week ago. But yes, star scheme's work the best in PBI.

Also dont forget a date table (dim) 😉

EDIT:
Besides that, if you really have alot of columns you can manage them and put them in maps. I allways do that with measure and dates and that cleans up alot. example:
Knipsel.PNG Knipsel2.PNG


Quality over Quantity


Did I answer your question? Mark my post as a solution!


Many thanks for your help.

Helpful resources

Announcements
April AMA free

Microsoft Fabric AMA Livestream

Join us Tuesday, April 09, 9:00 – 10:00 AM PST for a live, expert-led Q&A session on all things Microsoft Fabric!

March Fabric Community Update

Fabric Community Update - March 2024

Find out what's new and trending in the Fabric Community.